# US 2021/0241004 A1 · Object fence generation for lane assignment in autonomous machine applications > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0241004 A1 - Document ID: US20210241004A1 - Status: Published application - Kind: A1 - Application number: 17/234,475 - Filed: 2021-04-19 - Published: 2021-08-05 - Earliest claimed: 2019-08-08 - Assignees: NVIDIA Corporation - Applicants: NVIDIA Corporation - Inventors: Josh Abbott; Miguel Sainz Serra; Zhaoting Ye; David Nister - Web page: https://patentgrep.com/patent/US20210241004A1 ## CPC - G06T 7/73: using feature-based methods - G06V 20/588: {Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road} - G06T 7/12: Edge-based segmentation - G06T 7/70: Determining position or orientation of objects or cameras (camera calibration G06T7/80) - G06T 11/00: Two-dimensional [2D] image generation - G06T 11/20: Drawing from basic elements - G06V 20/58: Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - G06T 2207/20132: Image cropping - G06T 2207/30256: Lane; Road marking - G06T 2207/30261: Obstacle - G06T 2210/12: Bounding box - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract In various examples, object fence corresponding to objects detected by an ego-vehicle may be used to determine overlap of the object fences with lanes on a driving surface. A lane mask may be generated corresponding to the lanes on the driving surface, and the object fences may be compared to the lanes of the lane mask to determine the overlap. Where an object fence is located in more than one lane, a boundary scoring approach may be used to determine a ratio of overlap of the boundary fence, and thus the object, with each of the lanes. The overlap with one or more lanes for each object may be used to determine lane assignments for the objects, and the lane assignments may be used by the ego-vehicle to determine a path or trajectory along the driving surface. ## Claims ### Claim 1 A processor comprising: processing circuitry to generate an object fence based at least in part on cropping out a portion of a bounding shape associated with an object detected in an environment of an ego machine, and to execute one or more operations for the ego machine using the object fence. ### Claim 2 (depends on 1) The processor of claim 1, wherein the portion of the bounding shape is determined based at least in part on determined drivable freespace within the environment. ### Claim 3 (depends on 1) The processor of claim 1, wherein the portion of the bounding shape includes an upper portion of the bounding shape, and the object fence corresponds to a lower portion of the bounding shape. ### Claim 4 (depends on 1) The processor of claim 1, wherein the one or more operations include at least one of: assigning the object to a lane using the object fence, determining a path or trajectory through at least a portion of the environment, performing obstacle avoidance, or updating a world model. ### Claim 5 (depends on 4) The processor of claim 4, wherein the object fence is generated in image space, and the assigning the object to the lane includes comparing one or more first pixels associated with the object fence to one or more second pixels determined to be associated with the lane. ### Claim 6 (depends on 4) The processor of claim 4, wherein the assigning the object to the lane includes comparing the object fence to a lane mask. ### Claim 7 (depends on 4) The processor of claim 4, wherein the assigning the object to the lane includes: determining one or more distances between one or more pairs of vertices of the object fence; calculating a sum of the one or more distances; and assigning the object to the lane based at least in part on the sum of the one or more distances. ### Claim 8 (depends on 1) The processor of claim 1, wherein the one or more operations include comparing a location of the object fence to one or more locations of one or more features in the environment. ### Claim 9 A system comprising: one or more sensors; one or more processing units; and one or more memory units storing instructions thereon that, when executed using the one or more processing units, cause the one or more processing units to execute operations comprising: receiving sensor data generated using the one or more sensors; applying the sensor data to one or more machine learning models; generating, using the one or more machine learning models and based at least in part on the sensor data, data indicative of an object fence location of an object in an environment; and performing one or more operations using the object fence location. ### Claim 10 (depends on 9) The system of claim 9, wherein the object fence location is computed in a world-space coordinate system, and the one or more operations further comprise: converting the object fence location to an image-space coordinate system, wherein the performing the one or more operations is executed using the object fence location in the image-space coordinate system. ### Claim 11 (depends on 9) The system of claim 9, wherein the one or more machine learning models include a deep neural network (DNN). ### Claim 12 (depends on 9) The system of claim 9, wherein the one or more operations include at least one of: assigning the object to a lane using the object fence location, determining a path or trajectory through at least a portion of the environment, performing obstacle avoidance, or updating a world model. ### Claim 13 (depends on 12) The system of claim 12, wherein the assigning the object to the lane includes comparing first pixels associated with the object fence location to second pixels determined to be associated with the lane. ### Claim 14 (depends on 9) The system of claim 9, wherein the system is comprised in at least one of: a control system for an autonomous or semi-autonomous machine; a perception system for an autonomous or semi-autonomous machine; a system for performing simulation operations; a system for performing deep learning operations; a system implemented using an edge device; a system implemented using a robot; a system incorporating one or more virtual machines (VMs); a system implemented at least partially in a data center; or a system implemented at least partially using cloud computing resources. ### Claim 15 A method comprising: generating an object fence from a computed bounding shape corresponding to an object in an environment of an ego machine; determining a location for the object within a representation of one or more lanes in the environment using the object fence. ### Claim 16 (depends on 15) The method of claim 15, wherein the object fence corresponds to a portion of the computed bounding shape after cropping the computed bounding shape using a drivable freespace boundary. ### Claim 17 (depends on 15) The method of claim 15, wherein the object fence corresponds to a lower portion of the computed bounding shape after cropping an upper portion of the bounding shape. ### Claim 18 (depends on 15) The method of claim 15, wherein the representation of the one or more lanes includes a lane mask generated using map information or one or more outputs of one or more deep neural networks (DNNs). ### Claim 19 (depends on 15) The method of claim 15, wherein the object fence is generated in image space, and the determining the location for the object includes comparing first pixels associated with the object fence to second pixels determined to be associated with the one or more lanes. ### Claim 20 (depends on 15) The method of claim 15, wherein the representation of the one or more lanes is generated using triangulation to connect, in a crisscross pattern, vertices between lane lines corresponding to a same lane of the one or more lanes.